Intratumoral Heterogeneity of the Epigenome Tali Mazor, Aleksandr Pankov, Jun S. Song, Joseph F. Costello  Cancer Cell  Volume 29, Issue 4, Pages 440-451 (April 2016) DOI: 10.1016/j.ccell.2016.03.009 Copyright © 2016 Elsevier Inc. Terms and Conditions

Figure 1 Inferring the Evolutionary History of a Tumor from Intratumoral Heterogeneity (A) Analysis of spatially distinct biopsies can be used to build a phylogeny that represents the evolutionary history of a tumor. Phylogeny tree branches are colored according to the contributions of each cell population: the black branch contains alterations that are shared among all biopsies, the red/blue branch contains alterations shared between the red and blue biopsies, and the red, blue, and purple branches represent those alterations uniquely present in a single biopsy. (B) Phylogenies have traditionally been built from genetic alterations, including somatic mutations and copy-number alterations. Phyloepigenetic trees have been built from genome-wide DNA methylation data and could be similarly derived from other epigenetic marks, including histone modification patterns, open chromatin, or RNA expression levels. Representative patterns of somatic mutations, copy-number alterations, and DNA methylation (black, methylated CpG site; yellow, unmethylated CpG site) are shown. To date, tumor histories derived from genetic alterations and DNA methylation show similar evolutionary patterns, raising the question of the extent to which these alterations are functionally related. Further work is still required to determine whether phylogenies derived from histone modifications and RNA expression also reflect similar evolutionary histories. Cancer Cell 2016 29, 440-451DOI: (10.1016/j.ccell.2016.03.009) Copyright © 2016 Elsevier Inc. Terms and Conditions

Figure 2 Clinical Implications of Intratumoral Heterogeneity (A) Low levels of intratumoral heterogeneity in DNA methylation (mITH) have been associated with improved progression-free survival (PFS) and overall survival (OS). (B) Two different timeline models of the relationship between PFS and mITH. A tumor initiates (far left) and expands while acquiring a series of epigenetic alterations, leading to three distinct subclones (shades of blue) at the time of surgical resection (vertical black line). In both scenarios, identical subclones are present at initial resection. Surgical resection removes the majority of tumor cells, although a small number remain which continue to evolve and eventually develop into the recurrence after a short (top) or long (bottom) period of progression-free survival (PFS). Here, we question whether the duration of PFS may correlate with the levels of mITH in the recurrent tumor (cell populations on right). (C) The impact of therapy on mITH is understudied. For a given number of subclones at diagnosis (top), therapy may selectively kill some subclones, leading to decreased mITH at recurrence (left), or it may not alter the diversity of subclones, leading to a recurrence with similar subclones as the diagnostic tumor (center), or it may promote increased mITH and expansion of novel subclones, leading to a more heterogeneous recurrence (right). Cancer Cell 2016 29, 440-451DOI: (10.1016/j.ccell.2016.03.009) Copyright © 2016 Elsevier Inc. Terms and Conditions